MATERIALS TRANSACTIONS
Online ISSN : 1347-5320
Print ISSN : 1345-9678
ISSN-L : 1345-9678
Special Issue on Superfunctional Nanomaterials by Severe Plastic Deformation
Basic Research on Multi-Directional Forging of AZ80Mg Alloy for Fabrication of Bulky Mechanical Components
Hiromi MiuraWataru NakamuraChihiro Watanabe
Author information
JOURNAL FREE ACCESS FULL-TEXT HTML

2023 Volume 64 Issue 7 Pages 1504-1514

Details
Abstract

In this study, a commercial hot-extruded AZ80Mg alloy was multi-directionally forged (MDFed) at room temperature by employing pass strains of Δε = 0.1. The effects of the combined processes of MDFing and ageing on the microstructural evolution and strengthening were precisely examined in advance. The coarse initial grains were gradually subdivided into ultrafine grains by multiple mechanical twinning and kinking. As observed, the multiple twinning effectively suppressed the evolution of the sharp basal texture and enabled MDFing at room temperature to high cumulative strains. Although the combined processes of MDFing and ageing tended to increase the hardness and yield stress compared to those fabricated using simple MDFing at lower cumulative strain regions, the mechanical properties were almost comparable and independent of the processes at regions of higher cumulative strain beyond ΣΔε = 2.0. Yield strength over 505 MPa, ultimate tensile strength of over 612 MPa and ductility of over 7% were constantly achieved in all the processes. Although certain selected processes were applied to bulk samples for fabricating the mechanical components, frequent cracking hindered the MDFing to high cumulative strain regions. This finding signified that adequate MDFing process is dependent on sample size. However, MDFing with smaller pass strains than Δε = 0.1 enabled MDFing to regions of high cumulative strain. Thus, bulk AZ80Mg alloy with well-balanced mechanical properties—yield strength of 420 MPa, ultimate tensile strength of 540 MPa, and ductility of 10%—could be successfully fabricated.

Fig. 11 Influence of tensile behavior on strain rate of AZ80Mg samples prepared by a process of MDFing to ΣΔε = 1.0, followed by ageing at 423 K for 3.6 ks and further MDFing to ΣΔε = 1.4, i.e., MDFing to ΣΔε = 2.4 in total.

1. Introduction

Considerable research has been conducted to improve the mechanical properties of Mg alloys and promote their industrial applications as structural materials. To this end, the formation of alloys with rare-earth elements is one of the most innovative methods, which can significantly increase the tensile strength up to 400 MPa by forming long-period stacking ordered (LPSO) phase.1) Although this LPSO phase enhances the thermal stability of the alloys, its brittle nature degrades the ductility. Despite the specific strengthening of Mg alloys with the addition of rare-earth elements, reducing their amount of addition is highly desirable because they have emerged as strategic materials.

On the other hand, grain refinement can improve the mechanical properties of Mg alloys, which has been carried out by conventional thermomechanical processes employing mechanisms of static and dynamic recrystallizations.2) However, the achieved grain size d is generally coarser than 1 µm because of the grain growth prevalent at elevated temperatures.3) Therefore, realzing a drastic improvement in the mechanical properties of Mg alloys is considerably challenging. Recently, certain methods involving severe plastic deformation (SPD) have been applied to Mg alloys to produce ultrafine-grained (UFGed) structures (d < 1 µm), with substantial increase in strength as well as ductility.46) Harai et al. performed high-pressure torsion at a temperature of 423 K and successfully fabricated UFGed AZ61Mg alloy with an average grain size of 0.2 µm and a relatively high hardness of 1.1 GPa.4) Miura et al. reported that UFGed AZ61Mg alloy produced by multi-directional forging (MDFing) under decreasing temperature conditions exhibited a superior balance of mechanical properties: ultimate tensile strength (UTS) of 550 MPa, ductility of approximately 10% at room temperature, and superplasticity of 680% elongation at elevated temperatures.5,6) Thus, the evolution of UFGed microstructure can drastically improve the mechanical properties of Mg alloys.

The SPD of Mg alloys appears as a challenging task to accomplish at room temperature owing to its limited number of activated slip planes, i.e., only basal one. Miura et al. developed a new technique for room-temperature MDFing, wherein small pass strains of Δε = 0.1 enabled MDFing to a region of high cumulative strain. They applied this technique to AZ80Mg alloy and achieved excellent mechanical properties of 650 MPa UTS and 9% plastic strain to fracture.7) More importantly, the mechanical properties were balanced through uniform UFG evolution and grain-orientation randomization induced by multiple twinning and suppression of the sharp basal texture evolution. To this end, certain studies have attempted to accomplish strengthening by ageing Mg alloys,1,5,8) as ageing improved the hot-extrusion ability as well as the mechanical properties.8) Although remarkable mechanical properties were attained at room-temperature MDFing, the effects of ageing or the combined processes of MDFing and ageing on the microstructural evolution and strengthening have yet not been revealed. It is assumed that ageing should also induce additional work hardening by the Orowan strengthening during MDFing as well as impediment of dislocation motion after MDFing to derive exceptional strengthening.

Nevertheless, the SPD methods are considered unsuitable for fabricating large bulky materials and industrial mass productions because of complicated shape-invariant batch-type processes. Among the various SPD methods, MDFing is potentially the most suitable technique for fabricating UFGed bulky metallic materials. In the present study, the various combined processes of room-temperature MDFing and ageing were applied to an AZ80Mg alloy, and the variations in the microstructure and mechanical properties were systematically investigated. Finally, certain selected processes were applied to large AZ80Mg alloy samples for fabricating high-strength and lightweight mechanical components for practical applications.

2. Experimental Procedure

In this study, a hot-extruded AZ80Mg alloy rod with an average grain size of 20 µm was discharge machined into small rectangular-shaped samples with dimensions of 22.2 × 21.2 × 20 mm3 (aspect ratio: 1.11:1.10:1.00). The chemical composition is presented in Table 1. This aspect ratio was employed for MDFing with pass strains of Δε = 0.1. As reported, such small pass strains can effectively suppress sharp (0001) texture evolution to enable MDFing to high cumulative strain regions at room temperature.7) The initial equiaxed microstructure with extremely few precipitations is displayed in Fig. 1(a) and (b). The samples were MDFed at room temperature on an Amsler universal mechanical testing machine at an initial strain rate of 3.0 × 10−3 s−1. The first forging axis was parallel to the extrusion axis, and MDFing was performed to a cumulative strain of ΣΔε = 2.9, i.e., 29 passes of forging at maximum. The forging axis during MDFing was changed by 90° from pass to pass.57) Thus, the samples were successfully forged up to 29 passes without any reshaping. The as-fabricated MDFed sample is displayed in Fig. 2(a). Prior to or during MDFing, certain samples were aged at 423 K for 1 h. This ageing condition follows the age-hardening behavior of the as-hot-extruded and MDFed AZ80Mg alloys, which are discussed later (Section 3.1). A relatively lower temperature and shorter ageing period were employed to avoid extensive recovery and recrystallization, because SPDed lightweight metallic materials are thermally unstable.9) Hereinafter, the samples processed by i) MDFing, ii) ageing followed by MDFing, and iii) MDFing followed by ageing and additional MDFing are referred to as MDF, age + MDF, and MDF + age + MDF, respectively.

Table 1 Chemical composition of the AZ80Mg alloy in mass%.
Fig. 1

Microstructures of the samples: (a) and (b) as-hot extruded and (c) aged at 473 K for 5.4 ks. Observation was carried out using (a) optical microscope and (b), (c) transmission electron microscopy, respectively.

Fig. 2

(a) Image of small-sized AZ80Mg alloy sample MDFed to cumulative strain of ΣΔε = 1.0; (b) schematic illustration depicting the preparation of tensile specimens and planes for microstructural observations. F.A. denotes final forging axis.

Before and after the processes of ageing and MDFing, the microstructures evolved on the plane parallel to the final forging axis (Fig. 2(b)) were observed using optical microscope, orientation-imaging microscopy (SEM-OIM/Hitachi S-4300 - TexSemLab. AnalySIS4), and transmission electron microscopy (TEM/Jeol 2000FX). The tensile samples with gauge dimensions of 2.5 × 5.0 × 0.7 mm3 were discharge machined from the previous forging plane such that the tensile direction was perpendicular to the final forging axis (Fig. 2(b)). The tensile tests were performed on an Instron-type mechanical testing machine at ambient temperature and at various initial strain rates from 1.0 × 10−5 s−1 to 1.0 × 10−1 s−1. Moreover, the variations in hardness during ageing and MDFing were examined using a micro-Vickers microhardness tester. In addition, the variation in the Young’s modulus was examined by employing the strain-gauge method and a dynamic ultramicroindenter (Shimadzu DUH-211).

Certain processes selected from these experiments with small samples were applied to substantially larger bulky samples with approximate dimensions of 142 × 141 × 128 mm3 using an enormous forging machine at Kawamoto Heavy Industries Ltd., Japan. Although the aspect ratio remained was maintained constant, i.e., 1.11:1.10:1.00, reshaping was not performed even in case of employing various pass strains to promote the yield. Furthermore, the tensile tests were performed following the same procedure as that employed for the small samples.

3. Results

3.1 MDFing of small-sized samples of AZ80Mg alloys with combined ageing treatment

The variations in hardness during ageing at 423 K and 473 K of the as-hot-extruded sample and MDFed samples to cumulative strains of ΣΔε = 1.0 and 2.0, respectively, are presented in Fig. 3. The ageing of the as-hot-extruded sample exhibited clear hardening of Δ123 HV at 423 K and Δ130 HV at 473 K at the peak hardness. In contrast, the softening was nearly consistent in the MDFed samples, irrespective of the MDFing and ageing conditions. Notably, the softening was more evident at higher temperatures and in samples MDFed to higher cumulative regions. Even in case of such insufficient age hardening, the TEM observations confirmed the dense precipitation (Fig. 1(c)). The softening observed during ageing in Fig. 3 demonstrated that softening by recovery surpassed the influence of age hardening. Consequently, an ageing condition of 423 K for 3.6 ks was selected for the combined processes of room-temperature MDFing and ageing.

Fig. 3

Ageing behavior at 423 K and 473 K of AZ80Mg alloys as-hot extruded (ΣΔε = 0) and MDFed in advance to ΣΔε = 1.0 and ΣΔε = 2.0. Samples are indicated by the combination of MDFing cumulative strain in advance and ageing temperature. “ΣΔε = 1.0, 473 K” shows that sample was MDFed to ΣΔε = 1.0 in advance and then aged at 473 K.

The true stress versus cumulative strain curves obtained from various combined processes of MDFing and ageing at 423 K for 3.6 ks of the small-sized samples are plotted in Fig. 4. In all cases, MDFing was successfully performed at room temperature up to high cumulative strains. Interestingly, all flow curves exhibited a clear yielding followed by work hardening. The peak flow stress at each pass of forging increased rapidly at the regions of low cumulative strain, and thereafter, gradually increased at the regions of intermediate and high cumulative strain. Notably, the peak flow stress did not decrease after ageing, thereby indicating the suitable ageing condition employed in Fig. 3. The oscillation of the peak flow stress at every three cycles, induced by the basal texture in the as-hot-extruded sample, tended to gradually decrease under simple MDFing to the region of high cumulative strain (Fig. 4(a)) owing to the dense formation of the mechanical twins that degrade the initial texture yielding the isotropic mechanical properties.7) However, the peak oscillations of flow stress was relatively more evident in the samples aged even after MDFing to high cumulative strain regions, because the precipitates would hinder mechanical twinning. Generally, much fewer twins were observed in the samples prepared by the combined processes of MDFing and ageing (refer to Fig. 6). Therefore, the grain-orientation randomization was relatively suppressed. In certain instances, surface cracking appeared at regions of high cumulative strain during MDFing of the aged samples, which was caused by the reduced plastic deformability and twinning in addition to the slightly higher work hardening in presence of precipitates. The hard β-phase of Mg17Al12 precipitates10) formed in the AZ series of the Mg alloys can suppress mechanical twinning.11) Observably, all flow stresses were saturated at ∼450 MPa, regardless of the processes.

Fig. 4

True stress vs. cumulative strain curves obtained by combined processes of MDFing of small samples at room temperature with pass strains Δε = 0.1 and ageing at 423 K for 3.6 ks. Red and black flow curves indicate MDFing before and after ageing respectively, i.e., (a) simply MDFing to ΣΔε = 2.0 (20 passes),7) (b) MDFing (5 passes) + ageing + MDFing (20 passes), (c) MDFing (10 passes) + ageing + MDFing (14 passes), (d) MDFing (21 passes) + ageing + MDFing (8 passes) and (e) ageing + MDFing (20 passes).

The variations in the hardness during the combined processes of MDFing and ageing at 423 K for 3.6 ks are summarized in Fig. 5. As observed in all the processes, the hardness monononically increased with the cumulative strain. Specifically, the process of ageing followed by MDFing exhibited the highest hardness. In case of ageing after MDFing to ΣΔε = 0.5, the hardness was almost comparable with that of simple MDFing. Although the hardness decreased considerably because of recovery (Fig. 3) in the sample processed by MDFing to ΣΔε = 1.0 followed by ageing, it became prominently higher after additional MDFing to high cumulative regions. Compared to the hardening in simple MDFing, that in the processes with ageing evidently indicated the influence of precipitates on work hardening during MDFing. Ultimately, the remarkably high hardness of over 1.3 GPa was achieved at ΣΔε = 2.0 in the combined processes of MDFing and ageing. The highest hardness of 1.36 GPa was observed in the process of MDFing to ΣΔε = 1.0, followed by ageing and further MDFing up to ΣΔε = 2.4. This is because of the more considerable work hardening under MDFing, despite the softening by ageing in advance. Thus, the samples prepared by the combined processes exhibited a higher hardness than those fabricated by simple MDFing at a high cumulative strain region. Moreover, the hardness of the AZ80Mg alloys prepared by the present processes of MDFing and ageing was extraordinarily higher than those of the AZ61Mg alloys fabricated by MDFing at room temperature12) and that under decreasing temperature conditions.5)

Fig. 5

Variations in the hardness of AZ80Mg alloys during combined processes of MDFing and ageing at 423 K for 3.6 ks. For comparison, the hardness achieved by the simple MDFing to ΣΔε = 2.0 is indicated by triangle symbols after Ref. 7). Results of MDFed AZ61Mg alloys prepared by MDFing at room temperature12) or MDFing under decreasing temperature conditions5) indicated by dotted lines for comparison.

3.2 Variations in microstructure during MDFing of small samples

The typical microstructures observed under TEM are displayed in Fig. 6. The coarse initial grains (Fig. 3) were drastically fragmented by the evolution of dense mechanical twins only after two passes of MDFing (Fig. 6(a)). As observed, the majority of the mechanical twins could be attributed to the {1 0 –1 2} tension, and in certain areas, the twins intersected each other. Upon further MDFing, the twins subdivided to form a much finer-grained structure (Fig. 6(b)). These characteristics of acicular UFGs varied from those of the equiaxed ones developed after MDFing at elevated temperatures.5,6) As measured by the line-intercept method, the average grain size at ΣΔε = 2.0 was 0.3 µm. Thus, the UFGed structure could be attained at a relatively low cumulative strain of ΣΔε = 2.0 owing to the heightened grain fragmentation under mechanical twinning. However, the grains attained by the combined processes of MDFing and ageing (Fig. 6(c)) appeared coarser than those obtained by simple MDFing (Fig. 6(b)) even to a higher cumulative region, thereby suggesting the suppression of twinning by precipitates.11) As depicted in Fig. 7, double twinning occurred frequently, and the twinning inside the mother twin grains contributed to sponteneous grain fragmentation (Fig. 7(b)). Moreover, the average twin spacing was ∼50 nm (Fig. 7(b)). Therefore, the actual average grain size should be smaller than 0.3 µm. Previous studies have reported that active double and multiple twinning—formed under high forging stresses over 400 MPa—significantly accelerates the grain fragmentation process.7) However, the twin characteristics cannot be easily detected in the crystallographical analysis of the mechanical twins owing to the large distortions induced during MDFing.

Fig. 6

Evolved microstructures in small samples by (a) MDFing to ΣΔε = 0.2, (b) MDFing to ΣΔε = 2.0 and (c) MDFing to ΣΔε = 1.0, followed by ageing at 423 K for 3.6 ks and further MDFing to ΣΔε = 2.4 in total (i.e., additional ΣΔε = 1.4). F.A. denotes final forging axis.

Fig. 7

Typical photographs of double twinning formed after MDFing to (a) ΣΔε = 0.2 and (b) ΣΔε = 2.0. F.A. indicates final forging axis.

The microstructural variations during MDFing were observed using OIM, and the typical inverse pole figure (IPF) maps and the corresponding inverse pole figures are illustrated in Fig. 8. Although the finer twins exhibit a high density of formation in Figs. 6 and 7, most of them were too small to be observed under OIM or could not be detected because of the misorientation variations and distorsion during MDFing. Nevertheless, as observed from Fig. 8, the coarse initial grains were gradually subdivided by the mechanical twins and kinks at the macroscopic scale. The variations in the distribution of the twin characteristics by MDFing observed under OIM are summarized in Table 2. Overall, the majority of the twins observed under OIM were {1 0 –1 2} tension, and 99.7% of the twins formed at ΣΔε = 0.1 and 92.1% of the twins formed at ΣΔε = 0.7. Although the {1 0 –1 1}, {1 1 –2 1}, and {1 1 –2 2} twins tended to increase with the cumulative strain, their fraction was considerably low and not clearly observed under OIM. Compared to other mechanical twins, the more prevalent occurrence of the tension twins at the regions of lower cumulative strain could be attributed to the lower value of the critical resolved shear stress (CRSS). However, the other mechanical twins much with higher CRSS increased at regions of higher cumulative strain. This was induced by the substantially higher forging stresses compared to those of the CRSSes (Fig. 4), which will be discussed later. Interestingly, the intensity of the basal texture in Fig. 8(b) and (c) initially increased but later decreased after three passes of forging on the same plane. The reduced intensity of the basal texture at regions of higher cumulative strain can be related to the high-density mechanical twins formed during MDFing with a small pass strains of ε = 0.1,7) which are discussed later.

Fig. 8

Variations in microstructure with increasing cumulative strain of MDFing: (a) as-hot extruded, (b) MDFed to ΣΔε = 0.1, and (c) MDFed to ΣΔε = 0.7. Notably, the same plane of the sample was observed (a) before forging, (b) after one-pass forging, and (c) after three passes of forging at the same plane. Corresponding inverse pole figures with the maximum intensities are presented. In (b), white arrows indicate examples of kinks. Characteristics of grain boundaries are denoted as follows: 1°–3° boundaries (white strip); 3°–15° boundaries (black strip), >15° boundaries (black-bold lines), and S3 twin boundaries (purple bold lines).

Table 2 Change in the twin character distribution by MDFing. Fraction here indicates the ratio among the all grain boundaries.

3.3 Variations in tensile properties by combined processes of MDFing and ageing

The AZ80Mg alloys prepared using various combined processes of MDFing and ageing were tensile tested at room temperature. The typical flow curves are plotted in Fig. 9. As observed, the simply MDFed AZ80Mg alloys exhibited a drastic increase in the tensile strength at lower cumulative strain regions, which became gradual at regions of highercumulative strain. Note that an extraordinarily high yield strength of 530 MPa and UTS of 650 MPa was obtained with a moderate ductility of 9% at ΣΔε = 2.0.7) In other combined processes, similar mechanical properties could be attained irrespective of the processing conditions. The mechanical properties of the AZ80Mg alloys achieved by various processes of MDFing and ageing are listed in Table 3. As observed, the hardness and tensile strength obtained in all the combined processes were remarkably high. The yield stress and UTS greater than 500 and 600 MPa, respectively, remained almost constant for all processes. In particular, the high yield stress can be attributed to the combined effects of strain hardening and UFG evolution. More importantly, the mechanical properties of the forged samples were exceptionally balanced with a ductility of over 7% and tensile strength greater than 600 MPa even after SPD. It is interesting to note that marginal variations were observed in the slope at the elastic regions of the flow curves (Fig. 9). Specifically, the slope appeared to increase with the cumulative stain, which were examined based on the dynamic-hardness test and strain-gauge method. Therefore, the measurements were potentially affected by MDFing, and the measured average value of the Young’s modulus increased gradually from 43 GPa of the as-hot-extruded sample to approximately 54 GPa of the samples MDFed to ΣΔε = 2.0. After MDFing to ΣΔε = 2.0, the anisotropy in the Young’s modulus was not evident. Therefore, the variations in the Young’s modulus could be affected by the grain size, density of defects, stored strain energy, and other related factors. Thus, these aspects are subject to future investigation.

Fig. 9

True stress vs. nominal strain curves obtained by tensile tests of various processed AZ80Mg samples at an initial strain rate of 1.0 × 10−3 s−1 at room temperature. Tensile tests were carried out to the normal direction of the final forging axis or parallel to the hot extrusion direction. The ageing conditions were (a) at 473 K for 5.4 ks, and the others at 423 K for 3.6 ks. The process of “ΣΔε = 1.0 + age + ΣΔε = 1.4” indicates that MDFing was carried out to ΣΔε = 1.0 first and, then, followed by ageing and further MDFing to ΣΔε = 1.4, i.e., MDFing to ΣΔε = 2.4 in total. Some of the flow curves in (a) are after Ref. 7).

Table 3 Mechanical properties of AZ80Mg alloys processed by simple MDFing and combined processes of MDFing and ageing at 423 K for 3.6 ks. Tensile test was conducted at a strain rate of 1.0 × 10−3 s−1 at room temperature. The process of “ΣΔε = 0.5 + age + ΣΔε = 2.0” indicates that MDFing was performed to ΣΔε = 0.5 in advance, followed by ageing and additional MDFing to ΣΔε = 2.0, i.e., MDFing to ΣΔε = 2.5 in total.

4. Discussion for Small Samples

Small-sized AZ80Mg alloy samples were successfully MDFed at room temperature to a maximum cumulative strain of ΣΔε = 2.9 by the combined processes of MDFing and ageing at 423 K for 3.6 ks. The coarse initial grains were primarily subdivided based on the mechanical twins forming the acicular UFGed structure. The average grain size achieved was ∼0.3 µm at minimum, and all the fabricated AZ80 alloys manifested ultrahigh strength greater than 600 MPa with moderate ductility, irrespective of the processes. MDFing at extremely high forging stress level involved various phenomena that never occurred under the usual conditions of the conventional plastic deformation of Mg alloys. That is, i) a forging stress of ∼450 MPa cannot be realized under conventional conditions, because failure occurs before attaining such high flow stress, ii) such extremely high flow stress is repeatedly applied with the changing forging axis during MDFing, and therefore, various kinds and types of highly dense mechanical twins are formed even at high critical shear stresses and small shear factors. The mechanisms of UFG evolution and the superior mechanical properties are discussed below.

4.1 Grain fragmentation to form UFGed structure

The coarse initial grains in a hot-extruded AZ80Mg alloy were subdivided primarily by multiple twinning and their intersection with each other during MDFing (Figs. 68) to form an acicular UFGed structure. In particular, this feature is completely different from that of the samples fabricated by MDFing at elevated temperatures, employing dynamic recrystallization that derives equiaxed UFGed microstructures.5,6) The twin width ranged from 0.5–4 µm at the early stage of simple MDFing and diminished to 100 nm or less at higher cumulative regions (Fig. 6(a), (b); Fig. 7(b)). The width of the twins and the twin-boundary spacing increased relatively based on the combined processes of MDFing and ageing compared to those of simple MDFing. After MDFing to ΣΔε = 2.4, it was wider than 1 µm (Fig. 6(c)) in the former cases. Based on a high-temperature torsion experiment, Barnett et al. have reported that mechanical twinning becomes considerably difficult when the grains are finer than 5 µm.13) Moreover, their report stated the existence of a critical grain size for mechanical twinning. Nonetheless, twins were densely formed even in a fine-grained structure by MDFing at ambient temperature. Presumably, this variation was caused by the deformation temperature and applied flow stress, i.e., in case of deformation at elevated temperatures, the applied flow stress diminishes considerably owing to the dynamic recrystallization, which induces flow softening.3) Thus, the restriction of the applied flow stress should eliminate the mechanical twinning. Conversely, considerable work hardening continues up to the high strain regions during MDFing at ambient temperature. An abnormally high forging stress beyond 400 MPa, which substantially exceeds the fracture stresses of Mg alloys at typical deformation conditions, can induce the dense formation of various types of mechanical twins and their intersections even in UFGed structure. Nevertheless, under an insufficiently low shear factor of the twinning planes, it may subsequently increase after altering the forging axes to enable twinning. Consequently, homogeneous UFGed structures with predominant fragmentations were developed by mechanical twinning. If the samples were aged before or during MDFing, the evolved twins were notably coarser and exhibited wider boundary spacing than those obtained with simple MDFing (Fig. 6), signifying that the fine precipitates suppress mechanical twinning. As such, this suppression of twinning by precipitates has been reported by Stanford and Barnett.14) Moreover, twinning can be effectively hindered by the presence of precipitates even at a high forging stress beyond 400 MPa.

Overall, various kinds of mechanical twins were formed during room-temperature MDFing. The most frequently observed twins were those located at {1 0 –1 2}, {1 1 –2 2}, and {1 0 –1 1}. For the remaining mechanical twins such as {1 1 –2 1} twins, they could be identified by the OIM analysis at a low probability, which imply the formation of other types of twins as well. The average CRSSs for {1 0 –1 2} extension and {1 0 –1 1} compression twinning have been estimated as 2 and 115 MPa, respectively.15) Consequently, the {1 0 –1 1} compression twins as well as the other rare twins, which were not frequent under normal deformation conditions, could appear due to the extraordinarily high applied forging stress greater than 400 MPa with the forging axis transforming with each pass of MDFing (refer to Fig. 4). Such high forging stress can contribute to double and multiple twining, and the aforementioned activations of various kinds of mechanical twins induced significant grain fragmentation and homogeneous acicular UFGed structures.7)

The evolution of the basal texture was effectively suppressed by employing small pass strains at Δε = 0.1 accompanied with dense double and multiple twinnings.7) The variation in the intensity of the basal texture during MDFing is depicted in Fig. 10. Although the intensity first increased rapidly, it decreased from 19.1 to 4.4 after displaying a peak at around ΣΔε = 0.1. Therefore, the intensity of the basal texture was notably weakened upon MDFing to regions with higher cumulative strain. The peak intensity of the basal texture at ΣΔε = 0.1 was almost equivalent to that obtained by simple compression until fracture. This finding signifies that the applied pass strains of Δε = 0.1 reflect the possible upper bound of continuous MDFing at regions of high cumulative strain.

Fig. 10

Variations in the intensity of basal texture during MDFing measured by OIM (refer to Fig. 7). Measurements were recorded thoroughly on the same plane. The intensity measured after simple compression to fracture is displayed for comparison.

The dense formation of the tensile twins, including large misorientations with respect to their mother grains, can cause grain-orientation randomization. Furthermore, the samples were rotated by 90° for the subsequent forging pass prior to the evolution of the sharp basal texture. By conducting a hot-compression experiment of AZ31 Mg alloy, Yang et al. demonstrated that the extent of crystal rotation required for the development of the basal texture increases rapidly if the applied nominal strain exceeds 10%.16) Consequently, the small pass strains of 0.1 (∼11%) applieid herein effectively contributed toward suppressing such large crystal rotations and evolution of sharp basal texture. Consequently, the combined effects of small pass strains and mechanical twinning during MDFing successfully reduced the evolution of the sharp basal texture. Accordingly, the supression of the texture evolution impacted the success of MDFing at room temperature to regions of high cumulative strain. Additionally, the reduction of the basal texture evolution effectively contributed toward excellent ductility, as discussed in the following section.

4.2 Mechanical properties

The CRSS for prismatic slip was one order of magnitude higher than that for basal slip.17) Overall, the activated slip system at ambient temperature essentially represented a basal slip. Therefore, nonbasal slip systems cannot contribute toward plastic deformation, except for specific conditions. As reported in prior research, intense basal texture developed by the dominant basal slips reduces ductility. Nevertheless, nonbasal slip systems can be activated under high forging stresses proximate to or beyond 400 MPa at room temperature during MDFing (Fig. 4). In addition, the extremely high forging stress influenced the work hardening caused by multiple slips. Note that the activation of nonbasal slip systems and multiple twinning can act as factors to suppress the evolution of sharp basal texture. The numerous slips along with distributed fine precipitates can induce large strain hardening. Furthermore, the work-hardening rate increases with the decreasing grain size.18) Thus, the accelerating effects of multiple slips combined with the increase in dislocation density and grain subdivision by mechanical twinning produces an extremely high hardness (Fig. 5).

The MDFed AZ80Mg alloys exhibited a superior balance of mechanical properties (Fig. 9 and Table 3): high yield stress and UTS of 530 and 650 MPa, respectively, with moderate ductility of 9% by simple MDFing and greater than 505 and 612 MPa, respectively, with a minimum ductility of 7% by the combined processes of ageing and MDFing. Miura et al. reported that a high yield stress and UTS of 300 and 430 MPa, respectively, can be achieved by performing MDFing of AZ61Mg alloy under decreasing temperature conditions.5) Their relatively lower yield stress of 300 MPa compared to the that recorded in the present study can be attributed to the recovery and dynamic recrystallization during MDFing at elevated temperatures. Alternatively, the rare-earth-added Mg alloys exhibited high yield stress proximate to 400 MPa.19) Therefore, the mechanical properties achieved herein were superior to those of the rare-earth-added Mg alloys and MDFed under decreasing temperature conditions. Notably, the AZ80Mg alloys MDFed at room temperature exhibited moderate ductility (>7%) even after SPD (Table 3). This ductility is considerably less than that of the samples fabricated using MDFing under decreasing temperature conditions but more prominent than those of the rare-earth-added Mg alloys. Therefore, an excellent balance of mechanical properties can be accomplished by MDFing a commercial AZ80Mg alloy. As depicted in Fig. 10, the intensity of the basal texture rapidly decreased with the increasing cumulative strain to induce orientation randomization. It is known that the poor ductility at room temperature of Mg alloys is caused by the limited slip systems activated and basal texture evolution. Nonetheless, moderate ductility can be attained even after SPD by MDFing. This must be because of i) orientation randomization to enable crystallographical rotation and ii) multiple slips caused by adequately high flow stress to activate nonbasal slip systems during the tensile test.

As reported, the room temperature grain-boundary sliding (GBS) enhances the ductility if the grain size becomes finer than approximately a few micrometers.5,6) Notably, the effect of GBS on ductility is more prominent if the tensile tests are conducted at low strain rates. To evaluate the influence of GBS on ductility, the MDFed samples were tensile tested at various strain rates, and the typical flow curves are illustrated in Fig. 11. As the stress–strain curves were independent of the strain rate, the variations were not evident as observed in the other UFGed Mg alloy MDFed at continually decreasing temperature conditions.5,6) The increase in strength and the reduction in ductility with the increasing strain rate is typical in tension and frequently observed in various metallic materials with normal-size grained structures.20,21) Therefore, the GBS poses negligible influence on the mechanical properties of AZ80Mg alloy MDFed at room temperature owing to the impediment of the GBS by the acicular microstructures, i.e., edge-shaped ultrafine grains. Extremely work-hardened grains might prevent plastic accommodation that is required for extensive GBS.

Fig. 11

Influence of tensile behavior on strain rate of AZ80Mg samples prepared by a process of MDFing to ΣΔε = 1.0, followed by ageing at 423 K for 3.6 ks and further MDFing to ΣΔε = 1.4, i.e., MDFing to ΣΔε = 2.4 in total.

Several reports insist that double or multiple twinning triggers cracking,22) and their conclusion is based on the observations of double twins at the fractured surfaces. However, the present results disprove the aforementioned hypothesis that double twinning initiates cracking. Presumably, the formation of double twins in the fractured area can be attributed to stress concentration. Thus, the present study demonstrates that a dense population of the mechanical twins does not induce cracking. In contrast, the present observations indicate that dense twins and double-twinning populations induce enormous benefits yielding superior mechanical properties.

5. Fabrication of Bulky-UFGed AZ80Mg Alloy and Fabrication of Bulky Mechanical Parts

Based on the systematic investigation of the microstructure and mechanical properties of small-sized AZ80Mg alloy samples prepared by the combined processes of ageing and MDFing, we can conclude that simple MDFing is the most efficient process for producing UFGed samples to manufacture mechanical components for practical applications. The methods and results are described below, but all the experimental conditions could not be specified here for brevity and because of classification. Although the most suitable process, i.e., simple MDFing with pass strains of Δε = 0.1, was applied, frequent occurrences of fracture were observed before MDFing to regions of high cumulative strain. However, employing the combined proceeses of ageing and MDFing yielded the least suitable results. The frequent fracture of bulk samples during MDF can be attributed to the population of defects by the distribution of the defect size within the material volume and the defect size is statistically dependent on the material volume.23) These experimental results revealed that the experimental conditions suitable for small-sized samples cannot be adopted in all cases of the industrial production of bulk samples. Generally, coarser β-phase particles were more densely distributed in the as-hot-extruded bulky sample (Fig. 12). Moreover, the grain structure was coarser and not homegeneous compared to that in the small-sized sample (Fig. 1). These differences were potentially induced by the variations in the hot-extrusion ratio and the cooling rate after hot extrusion. After trial and error, additional smaller pass strains were employed for MDFing. The flow curves recorded during MDFing for small pass strains of Δε = 0.07 are depicted in Fig. 13, which were obtained using small-sized sample for demonstrative purposes. As a characteristic in Fig. 13, the forging stress decreased substantially compared with those by pass strains of Δε = 0.1 (Fig. 4). Nevertheless, the flow stress increased to 300 MPa at a cumulative strain of ΣΔε = 0.7. The prominent yielding in each flow curves indicated that each pass of MFDing intensified the plastic deformation. Although the maximum forging stress was 300 MPa, it should be adequate for multiple twinning. Moreover, dense twins were developed after two passes of MDFing of a small sample at forging stress of 300 MPa (Fig. 6(a)).

Fig. 12

Microstructure evolved in the hot-extrude large AZ80Mg alloy sample.

Fig. 13

True stress vs. cumulative strain curves obtained by MDFing with pass strains Δε = 0.07. Flow curves attained by simple MDFing of small-sized sample as example.

The bulk sample produced by MDFing and the corresponding tensile test results are presented in Fig. 14. As observed, a sufficiently large sample could be successfully fabricated without any cracks, which exhibited uniform tensile strength independent of the portion of the sample. The yield stress and UTS were approximately 420 and 540 MPa, respectively, with a ductility of 10%. Therefore, MDFing and strengthening could be successfully accomplished even in bulk samples by employing smaller pass strains than Δε = 0.1. These results would suggest that varying conditions are required for SPD process depending on the sample size.

Fig. 14

(a) Produced large bulky AZ80Mg alloy sample by MDFing and (b) results of tensile tests. The numbers 1 and 2 in (b) indicate sampling parts of tensile specimens at the center and the corner of the rectangular-shaped large sample, respectively.

Certain samples of the mechanical components for practical applications machined from the large-sized AZ80Mg alloy sample are displayed in Fig. 15. These mechanical components can be employed as substitutes for duralumin and super duralumin products by leveraging the advantages of the properties of UFGed Mg alloys.

Fig. 15

Samples of mechanical components for practical applications, wheel hub of bicycle, bolt, and gears, machined from the bulk AZ80Mg alloy sample produced by MDFing. Samples produced by Kawamoto Heavy Industries Ltd., Japan.

6. Conclusions

This study fabricated ultrafine-grained (UFGed) and high strength AZ80Mg alloy by simple multi-directional forging (MDFing) at room temperature and combined processes of MDFing and ageing at 423 K. The MDFing of small-sized samples could be successively performed to attain a maximum cumulative strain of ΣΔε = 2.9. The initial grains were fragmented by multiple twinning and kinking during MDFing to form an acicular UFGed microstructure with an average grain size of ∼0.3 µm. Although a similar UFGed structure could be attained by the combined processes with ageing before or during MDFing, the average grain size was marginally larger than that attained with simple MDFing, thereby signifying that the grain subdivision by twinning was delayed by the presence of the precipitates caused by suppression of mechanical twinning. The mechanical properties of the as-fabricated samples were almost the independent of the processes, and overall, displayed an excellent balance with a yield strength of 505 MPa, UTS of 612 MPa, ductility of 7% for each minimum. The moderate ductility even after severe plastic deformation could be reasonably attributed to the grain-orientation randomization by multiple twinning and multiple slips activated under extraordinarily high loading stress during tensile tests.

Nevertheless, the MDFing process validated with experiments on small-sized samples were not applicable to large-sized samples in all cases. Thus, the MDFing process must be tuned depending on the sample size, and MDFing of large-sized samples were successively accomplished by employing smaller pass strains compared to those for the small-sized ones. Finally, bulk UFGed AZ80Mg alloy samples can be produced to manufacture mechanical components for practical applications.

Acknowledgments

The authors acknowledge the financial support provided by the Grant-in-Aid for Scientific Research (KAKENHI) Grant No. 20H00305 and the Light Metal Educational Foundation. The authors also appreciate the great helps given by Kawamoto Heavy Industries Ltd., Japan, for the fabrication of bulky samples.

REFERENCES
 
© 2023 The Japan Institute of Metals and Materials
feedback
Top